src/gromacs/mdlib/sim_util.c

   1 /* -*- mode: c; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4; c-file-style: "stroustrup"; -*-
   2  *
   3  *
   4  *                This source code is part of
   5  *
   6  *                 G   R   O   M   A   C   S
   7  *
   8  *          GROningen MAchine for Chemical Simulations
   9  *
  10  *                        VERSION 3.2.0
  11  * Written by David van der Spoel, Erik Lindahl, Berk Hess, and others.
  12  * Copyright (c) 1991-2000, University of Groningen, The Netherlands.
  13  * Copyright (c) 2001-2004, The GROMACS development team,
  14  * check out http://www.gromacs.org for more information.
  15
  16  * This program is free software; you can redistribute it and/or
  17  * modify it under the terms of the GNU General Public License
  18  * as published by the Free Software Foundation; either version 2
  19  * of the License, or (at your option) any later version.
  20  *
  21  * If you want to redistribute modifications, please consider that
  22  * scientific software is very special. Version control is crucial -
  23  * bugs must be traceable. We will be happy to consider code for
  24  * inclusion in the official distribution, but derived work must not
  25  * be called official GROMACS. Details are found in the README & COPYING
  26  * files - if they are missing, get the official version at www.gromacs.org.
  27  *
  28  * To help us fund GROMACS development, we humbly ask that you cite
  29  * the papers on the package - you can find them in the top README file.
  30  *
  31  * For more info, check our website at http://www.gromacs.org
  32  *
  33  * And Hey:
  34  * GROwing Monsters And Cloning Shrimps
  35  */
  36 #ifdef HAVE_CONFIG_H
  37 #include <config.h>
  38 #endif
  39
  40 #ifdef GMX_CRAY_XT3
  41 #include<catamount/dclock.h>
  42 #endif
  43
  44
  45 #include <stdio.h>
  46 #include <time.h>
  47 #ifdef HAVE_SYS_TIME_H
  48 #include <sys/time.h>
  49 #endif
  50 #include <math.h>
  51 #include "typedefs.h"
  52 #include "string2.h"
  53 #include "gmxfio.h"
  54 #include "smalloc.h"
  55 #include "names.h"
  56 #include "confio.h"
  57 #include "mvdata.h"
  58 #include "txtdump.h"
  59 #include "pbc.h"
  60 #include "chargegroup.h"
  61 #include "vec.h"
  62 #include <time.h>
  63 #include "nrnb.h"
  64 #include "mshift.h"
  65 #include "mdrun.h"
  66 #include "update.h"
  67 #include "physics.h"
  68 #include "main.h"
  69 #include "mdatoms.h"
  70 #include "force.h"
  71 #include "bondf.h"
  72 #include "pme.h"
  73 #include "disre.h"
  74 #include "orires.h"
  75 #include "network.h"
  76 #include "calcmu.h"
  77 #include "constr.h"
  78 #include "xvgr.h"
  79 #include "trnio.h"
  80 #include "xtcio.h"
  81 #include "copyrite.h"
  82 #include "pull_rotation.h"
  83 #include "gmx_random.h"
  84 #include "domdec.h"
  85 #include "partdec.h"
  86 #include "gmx_wallcycle.h"
  87 #include "genborn.h"
  88
  89 #ifdef GMX_LIB_MPI
  90 #include <mpi.h>
  91 #endif
  92 #ifdef GMX_THREAD_MPI
  93 #include "tmpi.h"
  94 #endif
  95
  96 #include "adress.h"
  97 #include "qmmm.h"
  98
  99 #if 0
 100 typedef struct gmx_timeprint {
 101
 102 } t_gmx_timeprint;
 103 #endif
 104
 105 /* Portable version of ctime_r implemented in src/gmxlib/string2.c, but we do not want it declared in public installed headers */
 106 char *
 107 gmx_ctime_r(const time_t *clock,char *buf, int n);
 108
 109
 110 double
 111 gmx_gettime()
 112 {
 113 #ifdef HAVE_GETTIMEOFDAY
 114         struct timeval t;
 115         double seconds;
 116
 117         gettimeofday(&t,NULL);
 118
 119         seconds = (double) t.tv_sec + 1e-6*(double)t.tv_usec;
 120
 121         return seconds;
 122 #else
 123         double  seconds;
 124
 125         seconds = time(NULL);
 126
 127         return seconds;
 128 #endif
 129 }
 130
 131
 132 #define difftime(end,start) ((double)(end)-(double)(start))
 133
 134 void print_time(FILE *out,gmx_runtime_t *runtime,gmx_large_int_t step,
 135                 t_inputrec *ir, t_commrec *cr)
 136 {
 137     time_t finish;
 138     char   timebuf[STRLEN];
 139     double dt;
 140     char buf[48];
 141
 142 #ifndef GMX_THREAD_MPI
 143     if (!PAR(cr))
 144 #endif
 145     {
 146         fprintf(out,"\r");
 147     }
 148     fprintf(out,"step %s",gmx_step_str(step,buf));
 149     if ((step >= ir->nstlist))
 150     {
 151         if ((ir->nstlist == 0) || ((step % ir->nstlist) == 0))
 152         {
 153             /* We have done a full cycle let's update time_per_step */
 154             runtime->last = gmx_gettime();
 155             dt = difftime(runtime->last,runtime->real);
 156             runtime->time_per_step = dt/(step - ir->init_step + 1);
 157         }
 158         dt = (ir->nsteps + ir->init_step - step)*runtime->time_per_step;
 159
 160         if (ir->nsteps >= 0)
 161         {
 162             if (dt >= 300)
 163             {
 164                 finish = (time_t) (runtime->last + dt);
 165                 gmx_ctime_r(&finish,timebuf,STRLEN);
 166                 sprintf(buf,"%s",timebuf);
 167                 buf[strlen(buf)-1]='\0';
 168                 fprintf(out,", will finish %s",buf);
 169             }
 170             else
 171                 fprintf(out,", remaining runtime: %5d s          ",(int)dt);
 172         }
 173         else
 174         {
 175             fprintf(out," performance: %.1f ns/day    ",
 176                     ir->delta_t/1000*24*60*60/runtime->time_per_step);
 177         }
 178     }
 179 #ifndef GMX_THREAD_MPI
 180     if (PAR(cr))
 181     {
 182         fprintf(out,"\n");
 183     }
 184 #endif
 185
 186     fflush(out);
 187 }
 188
 189 #ifdef NO_CLOCK
 190 #define clock() -1
 191 #endif
 192
 193 static double set_proctime(gmx_runtime_t *runtime)
 194 {
 195     double diff;
 196 #ifdef GMX_CRAY_XT3
 197     double prev;
 198
 199     prev = runtime->proc;
 200     runtime->proc = dclock();
 201
 202     diff = runtime->proc - prev;
 203 #else
 204     clock_t prev;
 205
 206     prev = runtime->proc;
 207     runtime->proc = clock();
 208
 209     diff = (double)(runtime->proc - prev)/(double)CLOCKS_PER_SEC;
 210 #endif
 211     if (diff < 0)
 212     {
 213         /* The counter has probably looped, ignore this data */
 214         diff = 0;
 215     }
 216
 217     return diff;
 218 }
 219
 220 void runtime_start(gmx_runtime_t *runtime)
 221 {
 222     runtime->real = gmx_gettime();
 223     runtime->proc          = 0;
 224     set_proctime(runtime);
 225     runtime->realtime      = 0;
 226     runtime->proctime      = 0;
 227     runtime->last          = 0;
 228     runtime->time_per_step = 0;
 229 }
 230
 231 void runtime_end(gmx_runtime_t *runtime)
 232 {
 233     double now;
 234
 235     now = gmx_gettime();
 236
 237     runtime->proctime += set_proctime(runtime);
 238     runtime->realtime  = now - runtime->real;
 239     runtime->real      = now;
 240 }
 241
 242 void runtime_upd_proc(gmx_runtime_t *runtime)
 243 {
 244     runtime->proctime += set_proctime(runtime);
 245 }
 246
 247 void print_date_and_time(FILE *fplog,int nodeid,const char *title,
 248                          const gmx_runtime_t *runtime)
 249 {
 250     int i;
 251     char timebuf[STRLEN];
 252     char time_string[STRLEN];
 253     time_t tmptime;
 254
 255     if (fplog)
 256     {
 257         if (runtime != NULL)
 258         {
 259             tmptime = (time_t) runtime->real;
 260             gmx_ctime_r(&tmptime,timebuf,STRLEN);
 261         }
 262         else
 263         {
 264             tmptime = (time_t) gmx_gettime();
 265             gmx_ctime_r(&tmptime,timebuf,STRLEN);
 266         }
 267         for(i=0; timebuf[i]>=' '; i++)
 268         {
 269             time_string[i]=timebuf[i];
 270         }
 271         time_string[i]='\0';
 272
 273         fprintf(fplog,"%s on node %d %s\n",title,nodeid,time_string);
 274     }
 275 }
 276
 277 static void sum_forces(int start,int end,rvec f[],rvec flr[])
 278 {
 279   int i;
 280
 281   if (gmx_debug_at) {
 282     pr_rvecs(debug,0,"fsr",f+start,end-start);
 283     pr_rvecs(debug,0,"flr",flr+start,end-start);
 284   }
 285   for(i=start; (i<end); i++)
 286     rvec_inc(f[i],flr[i]);
 287 }
 288
 289 /*
 290  * calc_f_el calculates forces due to an electric field.
 291  *
 292  * force is kJ mol^-1 nm^-1 = e * kJ mol^-1 nm^-1 / e
 293  *
 294  * Et[] contains the parameters for the time dependent
 295  * part of the field (not yet used).
 296  * Ex[] contains the parameters for
 297  * the spatial dependent part of the field. You can have cool periodic
 298  * fields in principle, but only a constant field is supported
 299  * now.
 300  * The function should return the energy due to the electric field
 301  * (if any) but for now returns 0.
 302  *
 303  * WARNING:
 304  * There can be problems with the virial.
 305  * Since the field is not self-consistent this is unavoidable.
 306  * For neutral molecules the virial is correct within this approximation.
 307  * For neutral systems with many charged molecules the error is small.
 308  * But for systems with a net charge or a few charged molecules
 309  * the error can be significant when the field is high.
 310  * Solution: implement a self-consitent electric field into PME.
 311  */
 312 static void calc_f_el(FILE *fp,int  start,int homenr,
 313                       real charge[],rvec x[],rvec f[],
 314                       t_cosines Ex[],t_cosines Et[],double t)
 315 {
 316     rvec Ext;
 317     real t0;
 318     int  i,m;
 319
 320     for(m=0; (m<DIM); m++)
 321     {
 322         if (Et[m].n > 0)
 323         {
 324             if (Et[m].n == 3)
 325             {
 326                 t0 = Et[m].a[1];
 327                 Ext[m] = cos(Et[m].a[0]*(t-t0))*exp(-sqr(t-t0)/(2.0*sqr(Et[m].a[2])));
 328             }
 329             else
 330             {
 331                 Ext[m] = cos(Et[m].a[0]*t);
 332             }
 333         }
 334         else
 335         {
 336             Ext[m] = 1.0;
 337         }
 338         if (Ex[m].n > 0)
 339         {
 340             /* Convert the field strength from V/nm to MD-units */
 341             Ext[m] *= Ex[m].a[0]*FIELDFAC;
 342             for(i=start; (i<start+homenr); i++)
 343                 f[i][m] += charge[i]*Ext[m];
 344         }
 345         else
 346         {
 347             Ext[m] = 0;
 348         }
 349     }
 350     if (fp != NULL)
 351     {
 352         fprintf(fp,"%10g  %10g  %10g  %10g #FIELD\n",t,
 353                 Ext[XX]/FIELDFAC,Ext[YY]/FIELDFAC,Ext[ZZ]/FIELDFAC);
 354     }
 355 }
 356
 357 static void calc_virial(FILE *fplog,int start,int homenr,rvec x[],rvec f[],
 358                         tensor vir_part,t_graph *graph,matrix box,
 359                         t_nrnb *nrnb,const t_forcerec *fr,int ePBC)
 360 {
 361   int i,j;
 362   tensor virtest;
 363
 364   /* The short-range virial from surrounding boxes */
 365   clear_mat(vir_part);
 366   calc_vir(fplog,SHIFTS,fr->shift_vec,fr->fshift,vir_part,ePBC==epbcSCREW,box);
 367   inc_nrnb(nrnb,eNR_VIRIAL,SHIFTS);
 368
 369   /* Calculate partial virial, for local atoms only, based on short range.
 370    * Total virial is computed in global_stat, called from do_md
 371    */
 372   f_calc_vir(fplog,start,start+homenr,x,f,vir_part,graph,box);
 373   inc_nrnb(nrnb,eNR_VIRIAL,homenr);
 374
 375   /* Add position restraint contribution */
 376   for(i=0; i<DIM; i++) {
 377     vir_part[i][i] += fr->vir_diag_posres[i];
 378   }
 379
 380   /* Add wall contribution */
 381   for(i=0; i<DIM; i++) {
 382     vir_part[i][ZZ] += fr->vir_wall_z[i];
 383   }
 384
 385   if (debug)
 386     pr_rvecs(debug,0,"vir_part",vir_part,DIM);
 387 }
 388
 389 static void print_large_forces(FILE *fp,t_mdatoms *md,t_commrec *cr,
 390                                gmx_large_int_t step,real pforce,rvec *x,rvec *f)
 391 {
 392   int  i;
 393   real pf2,fn2;
 394   char buf[STEPSTRSIZE];
 395
 396   pf2 = sqr(pforce);
 397   for(i=md->start; i<md->start+md->homenr; i++) {
 398     fn2 = norm2(f[i]);
 399     /* We also catch NAN, if the compiler does not optimize this away. */
 400     if (fn2 >= pf2 || fn2 != fn2) {
 401       fprintf(fp,"step %s  atom %6d  x %8.3f %8.3f %8.3f  force %12.5e\n",
 402               gmx_step_str(step,buf),
 403               ddglatnr(cr->dd,i),x[i][XX],x[i][YY],x[i][ZZ],sqrt(fn2));
 404     }
 405   }
 406 }
 407
 408 void do_force(FILE *fplog,t_commrec *cr,
 409               t_inputrec *inputrec,
 410               gmx_large_int_t step,t_nrnb *nrnb,gmx_wallcycle_t wcycle,
 411               gmx_localtop_t *top,
 412               gmx_mtop_t *mtop,
 413               gmx_groups_t *groups,
 414               matrix box,rvec x[],history_t *hist,
 415               rvec f[],
 416               tensor vir_force,
 417               t_mdatoms *mdatoms,
 418               gmx_enerdata_t *enerd,t_fcdata *fcd,
 419               real *lambda,t_graph *graph,
 420               t_forcerec *fr,gmx_vsite_t *vsite,rvec mu_tot,
 421               double t,FILE *field,gmx_edsam_t ed,
 422               gmx_bool bBornRadii,
 423               int flags)
 424 {
 425     int    cg0,cg1,i,j;
 426     int    start,homenr;
 427     double mu[2*DIM];
 428     gmx_bool   bSepDVDL,bStateChanged,bNS,bFillGrid,bCalcCGCM,bBS;
 429     gmx_bool   bDoLongRange,bDoForces,bSepLRF;
 430     gmx_bool   bDoAdressWF;
 431     matrix boxs;
 432     real   e,v,dvdlambda[efptNR];
 433     real   dvdl_dum,lambda_dum;
 434     t_pbc  pbc;
 435     float  cycles_ppdpme,cycles_pme,cycles_seppme,cycles_force;
 436
 437     start  = mdatoms->start;
 438     homenr = mdatoms->homenr;
 439
 440     bSepDVDL = (fr->bSepDVDL && do_per_step(step,inputrec->nstlog));
 441
 442     clear_mat(vir_force);
 443
 444     if (PARTDECOMP(cr))
 445     {
 446         pd_cg_range(cr,&cg0,&cg1);
 447     }
 448     else
 449     {
 450         cg0 = 0;
 451         if (DOMAINDECOMP(cr))
 452         {
 453             cg1 = cr->dd->ncg_tot;
 454         }
 455         else
 456         {
 457             cg1 = top->cgs.nr;
 458         }
 459         if (fr->n_tpi > 0)
 460         {
 461             cg1--;
 462         }
 463     }
 464
 465     bStateChanged = (flags & GMX_FORCE_STATECHANGED);
 466     bNS           = (flags & GMX_FORCE_NS) && (fr->bAllvsAll==FALSE);
 467     bFillGrid     = (bNS && bStateChanged);
 468     bCalcCGCM     = (bFillGrid && !DOMAINDECOMP(cr));
 469     bDoLongRange  = (fr->bTwinRange && bNS && (flags & GMX_FORCE_DOLR));
 470     bDoForces     = (flags & GMX_FORCE_FORCES);
 471     bSepLRF       = (bDoLongRange && bDoForces && (flags & GMX_FORCE_SEPLRF));
 472     /* should probably move this to the forcerec since it doesn't change */
 473     bDoAdressWF   = ((fr->adress_type!=eAdressOff));
 474
 475     if (bStateChanged)
 476     {
 477         update_forcerec(fplog,fr,box);
 478
 479         /* Calculate total (local) dipole moment in a temporary common array.
 480          * This makes it possible to sum them over nodes faster.
 481          */
 482         calc_mu(start,homenr,
 483                 x,mdatoms->chargeA,mdatoms->chargeB,mdatoms->nChargePerturbed,
 484                 mu,mu+DIM);
 485     }
 486
 487   if (fr->ePBC != epbcNONE) {
 488     /* Compute shift vectors every step,
 489      * because of pressure coupling or box deformation!
 490      */
 491     if ((flags & GMX_FORCE_DYNAMICBOX) && bStateChanged)
 492       calc_shifts(box,fr->shift_vec);
 493
 494     if (bCalcCGCM) {
 495       put_charge_groups_in_box(fplog,cg0,cg1,fr->ePBC,box,
 496                                &(top->cgs),x,fr->cg_cm);
 497       inc_nrnb(nrnb,eNR_CGCM,homenr);
 498       inc_nrnb(nrnb,eNR_RESETX,cg1-cg0);
 499     }
 500     else if (EI_ENERGY_MINIMIZATION(inputrec->eI) && graph) {
 501       unshift_self(graph,box,x);
 502     }
 503   }
 504   else if (bCalcCGCM) {
 505     calc_cgcm(fplog,cg0,cg1,&(top->cgs),x,fr->cg_cm);
 506     inc_nrnb(nrnb,eNR_CGCM,homenr);
 507   }
 508
 509   if (bCalcCGCM) {
 510     if (PAR(cr)) {
 511       move_cgcm(fplog,cr,fr->cg_cm);
 512     }
 513     if (gmx_debug_at)
 514       pr_rvecs(debug,0,"cgcm",fr->cg_cm,top->cgs.nr);
 515   }
 516
 517 #ifdef GMX_MPI
 518   if (!(cr->duty & DUTY_PME)) {
 519     /* Send particle coordinates to the pme nodes.
 520      * Since this is only implemented for domain decomposition
 521      * and domain decomposition does not use the graph,
 522      * we do not need to worry about shifting.
 523      */
 524
 525     wallcycle_start(wcycle,ewcPP_PMESENDX);
 526
 527     bBS = (inputrec->nwall == 2);
 528     if (bBS) {
 529       copy_mat(box,boxs);
 530       svmul(inputrec->wall_ewald_zfac,boxs[ZZ],boxs[ZZ]);
 531     }
 532
 533     gmx_pme_send_x(cr,bBS ? boxs : box,x,
 534                    mdatoms->nChargePerturbed,lambda[efptCOUL],
 535                    ( flags & GMX_FORCE_VIRIAL),step);
 536
 537     wallcycle_stop(wcycle,ewcPP_PMESENDX);
 538   }
 539 #endif /* GMX_MPI */
 540
 541     /* Communicate coordinates and sum dipole if necessary */
 542     if (PAR(cr))
 543     {
 544         wallcycle_start(wcycle,ewcMOVEX);
 545         if (DOMAINDECOMP(cr))
 546         {
 547             dd_move_x(cr->dd,box,x);
 548         }
 549         else
 550         {
 551             move_x(fplog,cr,GMX_LEFT,GMX_RIGHT,x,nrnb);
 552         }
 553         /* When we don't need the total dipole we sum it in global_stat */
 554         if (bStateChanged && NEED_MUTOT(*inputrec))
 555         {
 556             gmx_sumd(2*DIM,mu,cr);
 557         }
 558         wallcycle_stop(wcycle,ewcMOVEX);
 559     }
 560     if (bStateChanged)
 561     {
 562
 563         /* update adress weight beforehand */
 564         if(bDoAdressWF)
 565         {
 566             /* need pbc for adress weight calculation with pbc_dx */
 567             set_pbc(&pbc,inputrec->ePBC,box);
 568             if(fr->adress_site == eAdressSITEcog)
 569             {
 570                 update_adress_weights_cog(top->idef.iparams,top->idef.il,x,fr,mdatoms,
 571                                           inputrec->ePBC==epbcNONE ? NULL : &pbc);
 572             }
 573             else if (fr->adress_site == eAdressSITEcom)
 574             {
 575                 update_adress_weights_com(fplog,cg0,cg1,&(top->cgs),x,fr,mdatoms,
 576                                           inputrec->ePBC==epbcNONE ? NULL : &pbc);
 577             }
 578             else if (fr->adress_site == eAdressSITEatomatom){
 579                 update_adress_weights_atom_per_atom(cg0,cg1,&(top->cgs),x,fr,mdatoms,
 580                                           inputrec->ePBC==epbcNONE ? NULL : &pbc);
 581             }
 582             else
 583             {
 584                 update_adress_weights_atom(cg0,cg1,&(top->cgs),x,fr,mdatoms,
 585                                            inputrec->ePBC==epbcNONE ? NULL : &pbc);
 586             }
 587         }
 588
 589         for(i=0; i<2; i++)
 590         {
 591             for(j=0;j<DIM;j++)
 592             {
 593                 fr->mu_tot[i][j] = mu[i*DIM + j];
 594             }
 595         }
 596     }
 597     if (fr->efep == efepNO)
 598     {
 599         copy_rvec(fr->mu_tot[0],mu_tot);
 600     }
 601     else
 602     {
 603         for(j=0; j<DIM; j++)
 604         {
 605             mu_tot[j] =
 606                 (1.0 - lambda[efptCOUL])*fr->mu_tot[0][j] + lambda[efptCOUL]*fr->mu_tot[1][j];
 607         }
 608     }
 609
 610     /* Reset energies */
 611     reset_enerdata(&(inputrec->opts),fr,bNS,enerd,MASTER(cr));
 612     clear_rvecs(SHIFTS,fr->fshift);
 613
 614     if (bNS)
 615     {
 616         wallcycle_start(wcycle,ewcNS);
 617
 618         if (graph && bStateChanged)
 619         {
 620             /* Calculate intramolecular shift vectors to make molecules whole */
 621             mk_mshift(fplog,graph,fr->ePBC,box,x);
 622         }
 623
 624         /* Reset long range forces if necessary */
 625         if (fr->bTwinRange)
 626         {
 627             /* Reset the (long-range) forces if necessary */
 628             clear_rvecs(fr->natoms_force_constr,bSepLRF ? fr->f_twin : f);
 629         }
 630
 631         /* Do the actual neighbour searching and if twin range electrostatics
 632          * also do the calculation of long range forces and energies.
 633          */
 634         for (i=0;i<efptNR;i++) {dvdlambda[i] = 0;}
 635         ns(fplog,fr,x,box,
 636            groups,&(inputrec->opts),top,mdatoms,
 637            cr,nrnb,lambda,dvdlambda,&enerd->grpp,bFillGrid,
 638            bDoLongRange,bDoForces,bSepLRF ? fr->f_twin : f);
 639         if (bSepDVDL)
 640         {
 641             fprintf(fplog,sepdvdlformat,"LR non-bonded",0.0,dvdlambda);
 642         }
 643         enerd->dvdl_lin[efptVDW] += dvdlambda[efptVDW];
 644         enerd->dvdl_lin[efptCOUL] += dvdlambda[efptCOUL];
 645
 646         wallcycle_stop(wcycle,ewcNS);
 647     }
 648
 649     if (inputrec->implicit_solvent && bNS)
 650     {
 651         make_gb_nblist(cr,inputrec->gb_algorithm,inputrec->rlist,
 652                        x,box,fr,&top->idef,graph,fr->born);
 653     }
 654
 655     if (DOMAINDECOMP(cr))
 656     {
 657         if (!(cr->duty & DUTY_PME))
 658         {
 659             wallcycle_start(wcycle,ewcPPDURINGPME);
 660             dd_force_flop_start(cr->dd,nrnb);
 661         }
 662     }
 663
 664     if (inputrec->bRot)
 665     {
 666         /* Enforced rotation has its own cycle counter that starts after the collective
 667          * coordinates have been communicated. It is added to ddCyclF to allow
 668          * for proper load-balancing */
 669         wallcycle_start(wcycle,ewcROT);
 670         do_rotation(cr,inputrec,box,x,t,step,wcycle,bNS);
 671         wallcycle_stop(wcycle,ewcROT);
 672     }
 673
 674     /* Start the force cycle counter.
 675      * This counter is stopped in do_forcelow_level.
 676      * No parallel communication should occur while this counter is running,
 677      * since that will interfere with the dynamic load balancing.
 678      */
 679     wallcycle_start(wcycle,ewcFORCE);
 680
 681     if (bDoForces)
 682     {
 683         /* Reset forces for which the virial is calculated separately:
 684          * PME/Ewald forces if necessary */
 685         if (fr->bF_NoVirSum)
 686         {
 687             if (flags & GMX_FORCE_VIRIAL)
 688             {
 689                 fr->f_novirsum = fr->f_novirsum_alloc;
 690                 if (fr->bDomDec)
 691                 {
 692                     clear_rvecs(fr->f_novirsum_n,fr->f_novirsum);
 693                 }
 694                 else
 695                 {
 696                     clear_rvecs(homenr,fr->f_novirsum+start);
 697                 }
 698             }
 699             else
 700             {
 701                 /* We are not calculating the pressure so we do not need
 702                  * a separate array for forces that do not contribute
 703                  * to the pressure.
 704                  */
 705                 fr->f_novirsum = f;
 706             }
 707         }
 708
 709         if (bSepLRF)
 710         {
 711             /* Add the long range forces to the short range forces */
 712             for(i=0; i<fr->natoms_force_constr; i++)
 713             {
 714                 copy_rvec(fr->f_twin[i],f[i]);
 715             }
 716         }
 717         else if (!(fr->bTwinRange && bNS))
 718         {
 719             /* Clear the short-range forces */
 720             clear_rvecs(fr->natoms_force_constr,f);
 721         }
 722
 723         clear_rvec(fr->vir_diag_posres);
 724     }
 725     if (inputrec->ePull == epullCONSTRAINT)
 726     {
 727         clear_pull_forces(inputrec->pull);
 728     }
 729
 730     /* update QMMMrec, if necessary */
 731     if(fr->bQMMM)
 732     {
 733         update_QMMMrec(cr,fr,x,mdatoms,box,top);
 734     }
 735
 736     if ((flags & GMX_FORCE_BONDED) && top->idef.il[F_POSRES].nr > 0)
 737     {
 738         /* Position restraints always require full pbc. Check if we already did it for Adress */
 739         if(!(bStateChanged && bDoAdressWF))
 740         {
 741             set_pbc(&pbc,inputrec->ePBC,box);
 742         }
 743         v = posres(top->idef.il[F_POSRES].nr,top->idef.il[F_POSRES].iatoms,
 744                    top->idef.iparams_posres,
 745                    (const rvec*)x,fr->f_novirsum,fr->vir_diag_posres,
 746                    inputrec->ePBC==epbcNONE ? NULL : &pbc,lambda[efptRESTRAINT],&(dvdlambda[efptRESTRAINT]),
 747                    fr->rc_scaling,fr->ePBC,fr->posres_com,fr->posres_comB);
 748         if (bSepDVDL)
 749         {
 750             fprintf(fplog,sepdvdlformat,
 751                     interaction_function[F_POSRES].longname,v,dvdlambda);
 752         }
 753         enerd->term[F_POSRES] += v;
 754         /* This linear lambda dependence assumption is only correct
 755          * when only k depends on lambda,
 756          * not when the reference position depends on lambda.
 757          * grompp checks for this.  (verify this is still the case?)
 758          */
 759         enerd->dvdl_nonlin[efptRESTRAINT] += dvdlambda[efptRESTRAINT]; /* if just the force constant changes, this is linear,
 760                                                                           but we can't be sure w/o additional checking that is
 761                                                                           hard to do at this level of code. Otherwise,
 762                                                                           the dvdl is not differentiable */
 763         inc_nrnb(nrnb,eNR_POSRES,top->idef.il[F_POSRES].nr/2);
 764         if ((inputrec->fepvals->n_lambda > 0) && (flags & GMX_FORCE_DHDL))
 765         {
 766             for(i=0; i<enerd->n_lambda; i++)
 767             {
 768                 lambda_dum = (i==0 ? lambda[efptRESTRAINT] : inputrec->fepvals->all_lambda[efptRESTRAINT][i-1]);
 769                 v = posres(top->idef.il[F_POSRES].nr,top->idef.il[F_POSRES].iatoms,
 770                            top->idef.iparams_posres,
 771                            (const rvec*)x,NULL,NULL,
 772                            inputrec->ePBC==epbcNONE ? NULL : &pbc,lambda_dum,&dvdl_dum,
 773                            fr->rc_scaling,fr->ePBC,fr->posres_com,fr->posres_comB);
 774                 enerd->enerpart_lambda[i] += v;
 775             }
 776         }
 777    }
 778
 779     if ((flags & GMX_FORCE_BONDED) && top->idef.il[F_FBPOSRES].nr > 0)
 780     {
 781         /* Flat-bottomed position restraints always require full pbc */
 782         if(!(bStateChanged && bDoAdressWF))
 783         {
 784             set_pbc(&pbc,inputrec->ePBC,box);
 785         }
 786         v = fbposres(top->idef.il[F_FBPOSRES].nr,top->idef.il[F_FBPOSRES].iatoms,
 787                      top->idef.iparams_fbposres,
 788                      (const rvec*)x,fr->f_novirsum,fr->vir_diag_posres,
 789                      inputrec->ePBC==epbcNONE ? NULL : &pbc,
 790                      fr->rc_scaling,fr->ePBC,fr->posres_com);
 791         enerd->term[F_FBPOSRES] += v;
 792         inc_nrnb(nrnb,eNR_FBPOSRES,top->idef.il[F_FBPOSRES].nr/2);
 793     }
 794
 795     /* Compute the bonded and non-bonded energies and optionally forces */
 796     do_force_lowlevel(fplog,step,fr,inputrec,&(top->idef),
 797                       cr,nrnb,wcycle,mdatoms,&(inputrec->opts),
 798                       x,hist,f,enerd,fcd,mtop,top,fr->born,
 799                       &(top->atomtypes),bBornRadii,box,
 800                       inputrec->fepvals,lambda,graph,&(top->excls),fr->mu_tot,
 801                       flags,&cycles_pme);
 802
 803     cycles_force = wallcycle_stop(wcycle,ewcFORCE);
 804
 805     if (ed)
 806     {
 807         do_flood(fplog,cr,x,f,ed,box,step,bNS);
 808     }
 809
 810     if (DOMAINDECOMP(cr))
 811     {
 812         dd_force_flop_stop(cr->dd,nrnb);
 813         if (wcycle)
 814         {
 815             dd_cycles_add(cr->dd,cycles_force-cycles_pme,ddCyclF);
 816         }
 817     }
 818
 819     if (bDoForces)
 820     {
 821         if (IR_ELEC_FIELD(*inputrec))
 822         {
 823             /* Compute forces due to electric field */
 824             calc_f_el(MASTER(cr) ? field : NULL,
 825                       start,homenr,mdatoms->chargeA,x,fr->f_novirsum,
 826                       inputrec->ex,inputrec->et,t);
 827         }
 828
 829         if (bDoAdressWF && fr->adress_icor == eAdressICThermoForce)
 830         {
 831             /* Compute thermodynamic force in hybrid AdResS region */
 832             adress_thermo_force(start,homenr,&(top->cgs),x,fr->f_novirsum,fr,mdatoms,
 833                                 inputrec->ePBC==epbcNONE ? NULL : &pbc);
 834         }
 835
 836         /* Communicate the forces */
 837         if (PAR(cr))
 838         {
 839             wallcycle_start(wcycle,ewcMOVEF);
 840             if (DOMAINDECOMP(cr))
 841             {
 842                 dd_move_f(cr->dd,f,fr->fshift);
 843                 /* Do we need to communicate the separate force array
 844                  * for terms that do not contribute to the single sum virial?
 845                  * Position restraints and electric fields do not introduce
 846                  * inter-cg forces, only full electrostatics methods do.
 847                  * When we do not calculate the virial, fr->f_novirsum = f,
 848                  * so we have already communicated these forces.
 849                  */
 850                 if (EEL_FULL(fr->eeltype) && cr->dd->n_intercg_excl &&
 851                     (flags & GMX_FORCE_VIRIAL))
 852                 {
 853                     dd_move_f(cr->dd,fr->f_novirsum,NULL);
 854                 }
 855                 if (bSepLRF)
 856                 {
 857                     /* We should not update the shift forces here,
 858                      * since f_twin is already included in f.
 859                      */
 860                     dd_move_f(cr->dd,fr->f_twin,NULL);
 861                 }
 862             }
 863             else
 864             {
 865                 pd_move_f(cr,f,nrnb);
 866                 if (bSepLRF)
 867                 {
 868                     pd_move_f(cr,fr->f_twin,nrnb);
 869                 }
 870             }
 871             wallcycle_stop(wcycle,ewcMOVEF);
 872         }
 873
 874         /* If we have NoVirSum forces, but we do not calculate the virial,
 875          * we sum fr->f_novirum=f later.
 876          */
 877         if (vsite && !(fr->bF_NoVirSum && !(flags & GMX_FORCE_VIRIAL)))
 878         {
 879             wallcycle_start(wcycle,ewcVSITESPREAD);
 880             spread_vsite_f(fplog,vsite,x,f,fr->fshift,FALSE,NULL,nrnb,
 881                            &top->idef,fr->ePBC,fr->bMolPBC,graph,box,cr);
 882             wallcycle_stop(wcycle,ewcVSITESPREAD);
 883
 884             if (bSepLRF)
 885             {
 886                 wallcycle_start(wcycle,ewcVSITESPREAD);
 887                 spread_vsite_f(fplog,vsite,x,fr->f_twin,NULL,FALSE,NULL,
 888                                nrnb,
 889                                &top->idef,fr->ePBC,fr->bMolPBC,graph,box,cr);
 890                 wallcycle_stop(wcycle,ewcVSITESPREAD);
 891             }
 892         }
 893
 894         if (flags & GMX_FORCE_VIRIAL)
 895         {
 896             /* Calculation of the virial must be done after vsites! */
 897             calc_virial(fplog,mdatoms->start,mdatoms->homenr,x,f,
 898                         vir_force,graph,box,nrnb,fr,inputrec->ePBC);
 899         }
 900     }
 901
 902     enerd->term[F_COM_PULL] = 0;
 903     if (inputrec->ePull == epullUMBRELLA || inputrec->ePull == epullCONST_F)
 904     {
 905         /* Calculate the center of mass forces, this requires communication,
 906          * which is why pull_potential is called close to other communication.
 907          * The virial contribution is calculated directly,
 908          * which is why we call pull_potential after calc_virial.
 909          */
 910         set_pbc(&pbc,inputrec->ePBC,box);
 911         dvdlambda[efptRESTRAINT] = 0;
 912         enerd->term[F_COM_PULL] +=
 913             pull_potential(inputrec->ePull,inputrec->pull,mdatoms,&pbc,
 914                            cr,t,lambda[efptRESTRAINT],x,f,vir_force,&(dvdlambda[efptRESTRAINT]));
 915         if (bSepDVDL)
 916         {
 917             fprintf(fplog,sepdvdlformat,"Com pull",enerd->term[F_COM_PULL],dvdlambda[efptRESTRAINT]);
 918         }
 919         enerd->dvdl_lin[efptRESTRAINT] += dvdlambda[efptRESTRAINT];
 920     }
 921
 922     /* Add the forces from enforced rotation potentials (if any) */
 923     if (inputrec->bRot)
 924     {
 925         wallcycle_start(wcycle,ewcROTadd);
 926         enerd->term[F_COM_PULL] += add_rot_forces(inputrec->rot, f, cr,step,t);
 927         wallcycle_stop(wcycle,ewcROTadd);
 928     }
 929
 930     if (PAR(cr) && !(cr->duty & DUTY_PME))
 931     {
 932         cycles_ppdpme = wallcycle_stop(wcycle,ewcPPDURINGPME);
 933         dd_cycles_add(cr->dd,cycles_ppdpme,ddCyclPPduringPME);
 934
 935         /* In case of node-splitting, the PP nodes receive the long-range
 936          * forces, virial and energy from the PME nodes here.
 937          */
 938         wallcycle_start(wcycle,ewcPP_PMEWAITRECVF);
 939         dvdlambda[efptCOUL] = 0;
 940         gmx_pme_receive_f(cr,fr->f_novirsum,fr->vir_el_recip,&e,&dvdlambda[efptCOUL],
 941                           &cycles_seppme);
 942         if (bSepDVDL)
 943         {
 944             fprintf(fplog,sepdvdlformat,"PME mesh",e,dvdlambda[efptCOUL]);
 945         }
 946         enerd->term[F_COUL_RECIP] += e;
 947         enerd->dvdl_lin[efptCOUL] += dvdlambda[efptCOUL];
 948         if (wcycle)
 949         {
 950             dd_cycles_add(cr->dd,cycles_seppme,ddCyclPME);
 951         }
 952         wallcycle_stop(wcycle,ewcPP_PMEWAITRECVF);
 953     }
 954
 955     if (bDoForces && fr->bF_NoVirSum)
 956     {
 957         if (vsite)
 958         {
 959             /* Spread the mesh force on virtual sites to the other particles...
 960              * This is parallellized. MPI communication is performed
 961              * if the constructing atoms aren't local.
 962              */
 963             wallcycle_start(wcycle,ewcVSITESPREAD);
 964             spread_vsite_f(fplog,vsite,x,fr->f_novirsum,NULL,
 965                            (flags & GMX_FORCE_VIRIAL),fr->vir_el_recip,
 966                            nrnb,
 967                            &top->idef,fr->ePBC,fr->bMolPBC,graph,box,cr);
 968             wallcycle_stop(wcycle,ewcVSITESPREAD);
 969         }
 970         if (flags & GMX_FORCE_VIRIAL)
 971         {
 972             /* Now add the forces, this is local */
 973             if (fr->bDomDec)
 974             {
 975                 sum_forces(0,fr->f_novirsum_n,f,fr->f_novirsum);
 976             }
 977             else
 978             {
 979                 sum_forces(start,start+homenr,f,fr->f_novirsum);
 980             }
 981             if (EEL_FULL(fr->eeltype))
 982             {
 983                 /* Add the mesh contribution to the virial */
 984                 m_add(vir_force,fr->vir_el_recip,vir_force);
 985             }
 986             if (debug)
 987             {
 988                 pr_rvecs(debug,0,"vir_force",vir_force,DIM);
 989             }
 990         }
 991     }
 992
 993     /* Sum the potential energy terms from group contributions */
 994     sum_epot(&(inputrec->opts),enerd);
 995
 996     if (fr->print_force >= 0 && bDoForces)
 997     {
 998         print_large_forces(stderr,mdatoms,cr,step,fr->print_force,x,f);
 999     }
1000 }
1001
1002 void do_constrain_first(FILE *fplog,gmx_constr_t constr,
1003                         t_inputrec *ir,t_mdatoms *md,
1004                         t_state *state,rvec *f,
1005                         t_graph *graph,t_commrec *cr,t_nrnb *nrnb,
1006                         t_forcerec *fr, gmx_localtop_t *top, tensor shake_vir)
1007 {
1008     int    i,m,start,end;
1009     gmx_large_int_t step;
1010     real   dt=ir->delta_t;
1011     real   dvdl_dum;
1012     rvec   *savex;
1013
1014     snew(savex,state->natoms);
1015
1016     start = md->start;
1017     end   = md->homenr + start;
1018
1019     if (debug)
1020         fprintf(debug,"vcm: start=%d, homenr=%d, end=%d\n",
1021                 start,md->homenr,end);
1022     /* Do a first constrain to reset particles... */
1023     step = ir->init_step;
1024     if (fplog)
1025     {
1026         char buf[STEPSTRSIZE];
1027         fprintf(fplog,"\nConstraining the starting coordinates (step %s)\n",
1028                 gmx_step_str(step,buf));
1029     }
1030     dvdl_dum = 0;
1031
1032     /* constrain the current position */
1033     constrain(NULL,TRUE,FALSE,constr,&(top->idef),
1034               ir,NULL,cr,step,0,md,
1035               state->x,state->x,NULL,
1036               state->box,state->lambda[efptBONDED],&dvdl_dum,
1037               NULL,NULL,nrnb,econqCoord,ir->epc==epcMTTK,state->veta,state->veta);
1038     if (EI_VV(ir->eI))
1039     {
1040         /* constrain the inital velocity, and save it */
1041         /* also may be useful if we need the ekin from the halfstep for velocity verlet */
1042         /* might not yet treat veta correctly */
1043         constrain(NULL,TRUE,FALSE,constr,&(top->idef),
1044                   ir,NULL,cr,step,0,md,
1045                   state->x,state->v,state->v,
1046                   state->box,state->lambda[efptBONDED],&dvdl_dum,
1047                   NULL,NULL,nrnb,econqVeloc,ir->epc==epcMTTK,state->veta,state->veta);
1048     }
1049     /* constrain the inital velocities at t-dt/2 */
1050     if (EI_STATE_VELOCITY(ir->eI) && ir->eI!=eiVV)
1051     {
1052         for(i=start; (i<end); i++)
1053         {
1054             for(m=0; (m<DIM); m++)
1055             {
1056                 /* Reverse the velocity */
1057                 state->v[i][m] = -state->v[i][m];
1058                 /* Store the position at t-dt in buf */
1059                 savex[i][m] = state->x[i][m] + dt*state->v[i][m];
1060             }
1061         }
1062     /* Shake the positions at t=-dt with the positions at t=0
1063      * as reference coordinates.
1064          */
1065         if (fplog)
1066         {
1067             char buf[STEPSTRSIZE];
1068             fprintf(fplog,"\nConstraining the coordinates at t0-dt (step %s)\n",
1069                     gmx_step_str(step,buf));
1070         }
1071         dvdl_dum = 0;
1072         constrain(NULL,TRUE,FALSE,constr,&(top->idef),
1073                   ir,NULL,cr,step,-1,md,
1074                   state->x,savex,NULL,
1075                   state->box,state->lambda[efptBONDED],&dvdl_dum,
1076                   state->v,NULL,nrnb,econqCoord,ir->epc==epcMTTK,state->veta,state->veta);
1077
1078         for(i=start; i<end; i++) {
1079             for(m=0; m<DIM; m++) {
1080                 /* Re-reverse the velocities */
1081                 state->v[i][m] = -state->v[i][m];
1082             }
1083         }
1084     }
1085     sfree(savex);
1086 }
1087
1088 void calc_enervirdiff(FILE *fplog,int eDispCorr,t_forcerec *fr)
1089 {
1090   double eners[2],virs[2],enersum,virsum,y0,f,g,h;
1091   double r0,r1,r,rc3,rc9,ea,eb,ec,pa,pb,pc,pd;
1092   double invscale,invscale2,invscale3;
1093   int    ri0,ri1,ri,i,offstart,offset;
1094   real   scale,*vdwtab;
1095
1096   fr->enershiftsix = 0;
1097   fr->enershifttwelve = 0;
1098   fr->enerdiffsix = 0;
1099   fr->enerdifftwelve = 0;
1100   fr->virdiffsix = 0;
1101   fr->virdifftwelve = 0;
1102
1103   if (eDispCorr != edispcNO) {
1104     for(i=0; i<2; i++) {
1105       eners[i] = 0;
1106       virs[i]  = 0;
1107     }
1108     if ((fr->vdwtype == evdwSWITCH) || (fr->vdwtype == evdwSHIFT)) {
1109       if (fr->rvdw_switch == 0)
1110         gmx_fatal(FARGS,
1111                   "With dispersion correction rvdw-switch can not be zero "
1112                   "for vdw-type = %s",evdw_names[fr->vdwtype]);
1113
1114       scale  = fr->nblists[0].tab.scale;
1115       vdwtab = fr->nblists[0].vdwtab;
1116
1117       /* Round the cut-offs to exact table values for precision */
1118       ri0 = floor(fr->rvdw_switch*scale);
1119       ri1 = ceil(fr->rvdw*scale);
1120       r0  = ri0/scale;
1121       r1  = ri1/scale;
1122       rc3 = r0*r0*r0;
1123       rc9  = rc3*rc3*rc3;
1124
1125       if (fr->vdwtype == evdwSHIFT) {
1126         /* Determine the constant energy shift below rvdw_switch */
1127         fr->enershiftsix    = (real)(-1.0/(rc3*rc3)) - vdwtab[8*ri0];
1128         fr->enershifttwelve = (real)( 1.0/(rc9*rc3)) - vdwtab[8*ri0 + 4];
1129       }
1130       /* Add the constant part from 0 to rvdw_switch.
1131        * This integration from 0 to rvdw_switch overcounts the number
1132        * of interactions by 1, as it also counts the self interaction.
1133        * We will correct for this later.
1134        */
1135       eners[0] += 4.0*M_PI*fr->enershiftsix*rc3/3.0;
1136       eners[1] += 4.0*M_PI*fr->enershifttwelve*rc3/3.0;
1137
1138       invscale = 1.0/(scale);
1139       invscale2 = invscale*invscale;
1140       invscale3 = invscale*invscale2;
1141
1142       /* following summation derived from cubic spline definition,
1143         Numerical Recipies in C, second edition, p. 113-116.  Exact
1144         for the cubic spline.  We first calculate the negative of
1145         the energy from rvdw to rvdw_switch, assuming that g(r)=1,
1146         and then add the more standard, abrupt cutoff correction to
1147         that result, yielding the long-range correction for a
1148         switched function.  We perform both the pressure and energy
1149         loops at the same time for simplicity, as the computational
1150         cost is low. */
1151
1152       for (i=0;i<2;i++) {
1153         enersum = 0.0; virsum = 0.0;
1154         if (i==0)
1155           offstart = 0;
1156         else
1157           offstart = 4;
1158         for (ri=ri0; ri<ri1; ri++) {
1159           r = ri*invscale;
1160           ea = invscale3;
1161           eb = 2.0*invscale2*r;
1162           ec = invscale*r*r;
1163
1164           pa = invscale3;
1165           pb = 3.0*invscale2*r;
1166           pc = 3.0*invscale*r*r;
1167           pd = r*r*r;
1168
1169           /* this "8" is from the packing in the vdwtab array - perhaps
1170             should be #define'ed? */
1171           offset = 8*ri + offstart;
1172           y0 = vdwtab[offset];
1173           f = vdwtab[offset+1];
1174           g = vdwtab[offset+2];
1175           h = vdwtab[offset+3];
1176
1177           enersum += y0*(ea/3 + eb/2 + ec) + f*(ea/4 + eb/3 + ec/2)+
1178             g*(ea/5 + eb/4 + ec/3) + h*(ea/6 + eb/5 + ec/4);
1179           virsum  +=  f*(pa/4 + pb/3 + pc/2 + pd) +
1180             2*g*(pa/5 + pb/4 + pc/3 + pd/2) + 3*h*(pa/6 + pb/5 + pc/4 + pd/3);
1181
1182         }
1183         enersum *= 4.0*M_PI;
1184         virsum  *= 4.0*M_PI;
1185         eners[i] -= enersum;
1186         virs[i]  -= virsum;
1187       }
1188
1189       /* now add the correction for rvdw_switch to infinity */
1190       eners[0] += -4.0*M_PI/(3.0*rc3);
1191       eners[1] +=  4.0*M_PI/(9.0*rc9);
1192       virs[0]  +=  8.0*M_PI/rc3;
1193       virs[1]  += -16.0*M_PI/(3.0*rc9);
1194     }
1195     else if ((fr->vdwtype == evdwCUT) || (fr->vdwtype == evdwUSER)) {
1196       if (fr->vdwtype == evdwUSER && fplog)
1197         fprintf(fplog,
1198                 "WARNING: using dispersion correction with user tables\n");
1199       rc3  = fr->rvdw*fr->rvdw*fr->rvdw;
1200       rc9  = rc3*rc3*rc3;
1201       eners[0] += -4.0*M_PI/(3.0*rc3);
1202       eners[1] +=  4.0*M_PI/(9.0*rc9);
1203       virs[0]  +=  8.0*M_PI/rc3;
1204       virs[1]  += -16.0*M_PI/(3.0*rc9);
1205     } else {
1206       gmx_fatal(FARGS,
1207                 "Dispersion correction is not implemented for vdw-type = %s",
1208                 evdw_names[fr->vdwtype]);
1209     }
1210     fr->enerdiffsix    = eners[0];
1211     fr->enerdifftwelve = eners[1];
1212     /* The 0.5 is due to the Gromacs definition of the virial */
1213     fr->virdiffsix     = 0.5*virs[0];
1214     fr->virdifftwelve  = 0.5*virs[1];
1215   }
1216 }
1217
1218 void calc_dispcorr(FILE *fplog,t_inputrec *ir,t_forcerec *fr,
1219                    gmx_large_int_t step,int natoms,
1220                    matrix box,real lambda,tensor pres,tensor virial,
1221                    real *prescorr, real *enercorr, real *dvdlcorr)
1222 {
1223     gmx_bool bCorrAll,bCorrPres;
1224     real dvdlambda,invvol,dens,ninter,avcsix,avctwelve,enerdiff,svir=0,spres=0;
1225     int  m;
1226
1227     *prescorr = 0;
1228     *enercorr = 0;
1229     *dvdlcorr = 0;
1230
1231     clear_mat(virial);
1232     clear_mat(pres);
1233
1234     if (ir->eDispCorr != edispcNO) {
1235         bCorrAll  = (ir->eDispCorr == edispcAllEner ||
1236                      ir->eDispCorr == edispcAllEnerPres);
1237         bCorrPres = (ir->eDispCorr == edispcEnerPres ||
1238                      ir->eDispCorr == edispcAllEnerPres);
1239
1240         invvol = 1/det(box);
1241         if (fr->n_tpi)
1242         {
1243             /* Only correct for the interactions with the inserted molecule */
1244             dens = (natoms - fr->n_tpi)*invvol;
1245             ninter = fr->n_tpi;
1246         }
1247         else
1248         {
1249             dens = natoms*invvol;
1250             ninter = 0.5*natoms;
1251         }
1252
1253         if (ir->efep == efepNO)
1254         {
1255             avcsix    = fr->avcsix[0];
1256             avctwelve = fr->avctwelve[0];
1257         }
1258         else
1259         {
1260             avcsix    = (1 - lambda)*fr->avcsix[0]    + lambda*fr->avcsix[1];
1261             avctwelve = (1 - lambda)*fr->avctwelve[0] + lambda*fr->avctwelve[1];
1262         }
1263
1264         enerdiff = ninter*(dens*fr->enerdiffsix - fr->enershiftsix);
1265         *enercorr += avcsix*enerdiff;
1266         dvdlambda = 0.0;
1267         if (ir->efep != efepNO)
1268         {
1269             dvdlambda += (fr->avcsix[1] - fr->avcsix[0])*enerdiff;
1270         }
1271         if (bCorrAll)
1272         {
1273             enerdiff = ninter*(dens*fr->enerdifftwelve - fr->enershifttwelve);
1274             *enercorr += avctwelve*enerdiff;
1275             if (fr->efep != efepNO)
1276             {
1277                 dvdlambda += (fr->avctwelve[1] - fr->avctwelve[0])*enerdiff;
1278             }
1279         }
1280
1281         if (bCorrPres)
1282         {
1283             svir = ninter*dens*avcsix*fr->virdiffsix/3.0;
1284             if (ir->eDispCorr == edispcAllEnerPres)
1285             {
1286                 svir += ninter*dens*avctwelve*fr->virdifftwelve/3.0;
1287             }
1288             /* The factor 2 is because of the Gromacs virial definition */
1289             spres = -2.0*invvol*svir*PRESFAC;
1290
1291             for(m=0; m<DIM; m++) {
1292                 virial[m][m] += svir;
1293                 pres[m][m] += spres;
1294             }
1295             *prescorr += spres;
1296         }
1297
1298         /* Can't currently control when it prints, for now, just print when degugging */
1299         if (debug)
1300         {
1301             if (bCorrAll) {
1302                 fprintf(debug,"Long Range LJ corr.: <C6> %10.4e, <C12> %10.4e\n",
1303                         avcsix,avctwelve);
1304             }
1305             if (bCorrPres)
1306             {
1307                 fprintf(debug,
1308                         "Long Range LJ corr.: Epot %10g, Pres: %10g, Vir: %10g\n",
1309                         *enercorr,spres,svir);
1310             }
1311             else
1312             {
1313                 fprintf(debug,"Long Range LJ corr.: Epot %10g\n",*enercorr);
1314             }
1315         }
1316
1317         if (fr->bSepDVDL && do_per_step(step,ir->nstlog))
1318         {
1319             fprintf(fplog,sepdvdlformat,"Dispersion correction",
1320                     *enercorr,dvdlambda);
1321         }
1322         if (fr->efep != efepNO)
1323         {
1324             *dvdlcorr += dvdlambda;
1325         }
1326     }
1327 }
1328
1329 void do_pbc_first(FILE *fplog,matrix box,t_forcerec *fr,
1330                   t_graph *graph,rvec x[])
1331 {
1332   if (fplog)
1333     fprintf(fplog,"Removing pbc first time\n");
1334   calc_shifts(box,fr->shift_vec);
1335   if (graph) {
1336     mk_mshift(fplog,graph,fr->ePBC,box,x);
1337     if (gmx_debug_at)
1338       p_graph(debug,"do_pbc_first 1",graph);
1339     shift_self(graph,box,x);
1340     /* By doing an extra mk_mshift the molecules that are broken
1341      * because they were e.g. imported from another software
1342      * will be made whole again. Such are the healing powers
1343      * of GROMACS.
1344      */
1345     mk_mshift(fplog,graph,fr->ePBC,box,x);
1346     if (gmx_debug_at)
1347       p_graph(debug,"do_pbc_first 2",graph);
1348   }
1349   if (fplog)
1350     fprintf(fplog,"Done rmpbc\n");
1351 }
1352
1353 static void low_do_pbc_mtop(FILE *fplog,int ePBC,matrix box,
1354                             gmx_mtop_t *mtop,rvec x[],
1355                             gmx_bool bFirst)
1356 {
1357   t_graph *graph;
1358   int mb,as,mol;
1359   gmx_molblock_t *molb;
1360
1361   if (bFirst && fplog)
1362     fprintf(fplog,"Removing pbc first time\n");
1363
1364   snew(graph,1);
1365   as = 0;
1366   for(mb=0; mb<mtop->nmolblock; mb++) {
1367     molb = &mtop->molblock[mb];
1368     if (molb->natoms_mol == 1 ||
1369         (!bFirst && mtop->moltype[molb->type].cgs.nr == 1)) {
1370       /* Just one atom or charge group in the molecule, no PBC required */
1371       as += molb->nmol*molb->natoms_mol;
1372     } else {
1373       /* Pass NULL iso fplog to avoid graph prints for each molecule type */
1374       mk_graph_ilist(NULL,mtop->moltype[molb->type].ilist,
1375                      0,molb->natoms_mol,FALSE,FALSE,graph);
1376
1377       for(mol=0; mol<molb->nmol; mol++) {
1378         mk_mshift(fplog,graph,ePBC,box,x+as);
1379
1380         shift_self(graph,box,x+as);
1381         /* The molecule is whole now.
1382          * We don't need the second mk_mshift call as in do_pbc_first,
1383          * since we no longer need this graph.
1384          */
1385
1386         as += molb->natoms_mol;
1387       }
1388       done_graph(graph);
1389     }
1390   }
1391   sfree(graph);
1392 }
1393
1394 void do_pbc_first_mtop(FILE *fplog,int ePBC,matrix box,
1395                        gmx_mtop_t *mtop,rvec x[])
1396 {
1397   low_do_pbc_mtop(fplog,ePBC,box,mtop,x,TRUE);
1398 }
1399
1400 void do_pbc_mtop(FILE *fplog,int ePBC,matrix box,
1401                  gmx_mtop_t *mtop,rvec x[])
1402 {
1403   low_do_pbc_mtop(fplog,ePBC,box,mtop,x,FALSE);
1404 }
1405
1406 void finish_run(FILE *fplog,t_commrec *cr,const char *confout,
1407                 t_inputrec *inputrec,
1408                 t_nrnb nrnb[],gmx_wallcycle_t wcycle,
1409                 gmx_runtime_t *runtime,
1410                 gmx_bool bWriteStat)
1411 {
1412   int    i,j;
1413   t_nrnb *nrnb_tot=NULL;
1414   real   delta_t;
1415   double nbfs,mflop;
1416   double cycles[ewcNR];
1417
1418   wallcycle_sum(cr,wcycle,cycles);
1419
1420   if (cr->nnodes > 1) {
1421     if (SIMMASTER(cr))
1422       snew(nrnb_tot,1);
1423 #ifdef GMX_MPI
1424     MPI_Reduce(nrnb->n,nrnb_tot->n,eNRNB,MPI_DOUBLE,MPI_SUM,
1425                MASTERRANK(cr),cr->mpi_comm_mysim);
1426 #endif
1427   } else {
1428     nrnb_tot = nrnb;
1429   }
1430
1431   if (SIMMASTER(cr)) {
1432     print_flop(fplog,nrnb_tot,&nbfs,&mflop);
1433     if (cr->nnodes > 1) {
1434       sfree(nrnb_tot);
1435     }
1436   }
1437
1438   if ((cr->duty & DUTY_PP) && DOMAINDECOMP(cr)) {
1439     print_dd_statistics(cr,inputrec,fplog);
1440   }
1441
1442 #ifdef GMX_MPI
1443     if (PARTDECOMP(cr))
1444     {
1445         if (MASTER(cr))
1446         {
1447             t_nrnb     *nrnb_all;
1448             int        s;
1449             MPI_Status stat;
1450
1451             snew(nrnb_all,cr->nnodes);
1452             nrnb_all[0] = *nrnb;
1453             for(s=1; s<cr->nnodes; s++)
1454             {
1455                 MPI_Recv(nrnb_all[s].n,eNRNB,MPI_DOUBLE,s,0,
1456                          cr->mpi_comm_mysim,&stat);
1457             }
1458             pr_load(fplog,cr,nrnb_all);
1459             sfree(nrnb_all);
1460         }
1461         else
1462         {
1463             MPI_Send(nrnb->n,eNRNB,MPI_DOUBLE,MASTERRANK(cr),0,
1464                      cr->mpi_comm_mysim);
1465         }
1466     }
1467 #endif
1468
1469   if (SIMMASTER(cr)) {
1470     wallcycle_print(fplog,cr->nnodes,cr->npmenodes,runtime->realtime,
1471                     wcycle,cycles);
1472
1473     if (EI_DYNAMICS(inputrec->eI)) {
1474       delta_t = inputrec->delta_t;
1475     } else {
1476       delta_t = 0;
1477     }
1478
1479     if (fplog) {
1480         print_perf(fplog,runtime->proctime,runtime->realtime,
1481                    cr->nnodes-cr->npmenodes,
1482                    runtime->nsteps_done,delta_t,nbfs,mflop);
1483     }
1484     if (bWriteStat) {
1485         print_perf(stderr,runtime->proctime,runtime->realtime,
1486                    cr->nnodes-cr->npmenodes,
1487                    runtime->nsteps_done,delta_t,nbfs,mflop);
1488     }
1489
1490     /*
1491     runtime=inputrec->nsteps*inputrec->delta_t;
1492     if (bWriteStat) {
1493       if (cr->nnodes == 1)
1494         fprintf(stderr,"\n\n");
1495       print_perf(stderr,nodetime,realtime,runtime,&ntot,
1496                  cr->nnodes-cr->npmenodes,FALSE);
1497     }
1498     wallcycle_print(fplog,cr->nnodes,cr->npmenodes,realtime,wcycle,cycles);
1499     print_perf(fplog,nodetime,realtime,runtime,&ntot,cr->nnodes-cr->npmenodes,
1500                TRUE);
1501     if (PARTDECOMP(cr))
1502       pr_load(fplog,cr,nrnb_all);
1503     if (cr->nnodes > 1)
1504       sfree(nrnb_all);
1505     */
1506   }
1507 }
1508
1509 extern void initialize_lambdas(FILE *fplog,t_inputrec *ir,int *fep_state,real *lambda,double *lam0)
1510 {
1511     /* this function works, but could probably use a logic rewrite to keep all the different
1512        types of efep straight. */
1513
1514     int i;
1515     t_lambda *fep = ir->fepvals;
1516
1517     if ((ir->efep==efepNO) && (ir->bSimTemp == FALSE)) {
1518         for (i=0;i<efptNR;i++)  {
1519             lambda[i] = 0.0;
1520             if (lam0)
1521             {
1522                 lam0[i] = 0.0;
1523             }
1524         }
1525         return;
1526     } else {
1527         *fep_state = fep->init_fep_state; /* this might overwrite the checkpoint
1528                                              if checkpoint is set -- a kludge is in for now
1529                                              to prevent this.*/
1530         for (i=0;i<efptNR;i++)
1531         {
1532             /* overwrite lambda state with init_lambda for now for backwards compatibility */
1533             if (fep->init_lambda>=0) /* if it's -1, it was never initializd */
1534             {
1535                 lambda[i] = fep->init_lambda;
1536                 if (lam0) {
1537                     lam0[i] = lambda[i];
1538                 }
1539             }
1540             else
1541             {
1542                 lambda[i] = fep->all_lambda[i][*fep_state];
1543                 if (lam0) {
1544                     lam0[i] = lambda[i];
1545                 }
1546             }
1547         }
1548         if (ir->bSimTemp) {
1549             /* need to rescale control temperatures to match current state */
1550             for (i=0;i<ir->opts.ngtc;i++) {
1551                 if (ir->opts.ref_t[i] > 0) {
1552                     ir->opts.ref_t[i] = ir->simtempvals->temperatures[*fep_state];
1553                 }
1554             }
1555         }
1556     }
1557
1558     /* Send to the log the information on the current lambdas */
1559     if (fplog != NULL)
1560     {
1561         fprintf(fplog,"Initial vector of lambda components:[ ");
1562         for (i=0;i<efptNR;i++)
1563         {
1564             fprintf(fplog,"%10.4f ",lambda[i]);
1565         }
1566         fprintf(fplog,"]\n");
1567     }
1568     return;
1569 }
1570
1571
1572 void init_md(FILE *fplog,
1573              t_commrec *cr,t_inputrec *ir,const output_env_t oenv,
1574              double *t,double *t0,
1575              real *lambda, int *fep_state, double *lam0,
1576              t_nrnb *nrnb,gmx_mtop_t *mtop,
1577              gmx_update_t *upd,
1578              int nfile,const t_filenm fnm[],
1579              gmx_mdoutf_t **outf,t_mdebin **mdebin,
1580              tensor force_vir,tensor shake_vir,rvec mu_tot,
1581              gmx_bool *bSimAnn,t_vcm **vcm, t_state *state, unsigned long Flags)
1582 {
1583     int  i,j,n;
1584     real tmpt,mod;
1585
1586     /* Initial values */
1587     *t = *t0       = ir->init_t;
1588
1589     *bSimAnn=FALSE;
1590     for(i=0;i<ir->opts.ngtc;i++)
1591     {
1592         /* set bSimAnn if any group is being annealed */
1593         if(ir->opts.annealing[i]!=eannNO)
1594         {
1595             *bSimAnn = TRUE;
1596         }
1597     }
1598     if (*bSimAnn)
1599     {
1600         update_annealing_target_temp(&(ir->opts),ir->init_t);
1601     }
1602
1603     /* Initialize lambda variables */
1604     initialize_lambdas(fplog,ir,fep_state,lambda,lam0);
1605
1606     if (upd)
1607     {
1608         *upd = init_update(fplog,ir);
1609     }
1610
1611
1612     if (vcm != NULL)
1613     {
1614         *vcm = init_vcm(fplog,&mtop->groups,ir);
1615     }
1616
1617     if (EI_DYNAMICS(ir->eI) && !(Flags & MD_APPENDFILES))
1618     {
1619         if (ir->etc == etcBERENDSEN)
1620         {
1621             please_cite(fplog,"Berendsen84a");
1622         }
1623         if (ir->etc == etcVRESCALE)
1624         {
1625             please_cite(fplog,"Bussi2007a");
1626         }
1627     }
1628
1629     init_nrnb(nrnb);
1630
1631     if (nfile != -1)
1632     {
1633         *outf = init_mdoutf(nfile,fnm,Flags,cr,ir,oenv);
1634
1635         *mdebin = init_mdebin((Flags & MD_APPENDFILES) ? NULL : (*outf)->fp_ene,
1636                               mtop,ir, (*outf)->fp_dhdl);
1637     }
1638
1639     if (ir->bAdress)
1640     {
1641       please_cite(fplog,"Fritsch12");
1642       please_cite(fplog,"Junghans10");
1643     }
1644     /* Initiate variables */
1645     clear_mat(force_vir);
1646     clear_mat(shake_vir);
1647     clear_rvec(mu_tot);
1648
1649     debug_gmx();
1650 }
1651
1652
1653
1654